Skip to main content

GPT-5.4 mini (Non-Reasoning)

OpenAIGPTProprietary

Description

GPT-5.4 mini is OpenAI's strongest mini model yet for coding, computer use, and subagents. It brings many of the strengths of GPT-5.4 to a faster, more efficient model designed for high-volume workloads, with strong coding, reasoning, multimodal understanding, and tool use.

Release Date
2026-03-17
Parameters
Context Length
400K
Modalities
file, image, text

Capability Radar

20
general
28
coding
61
reasoning
42
scienceest.
60
agents
85
multimodal

Science uses a reasoning proxy when dedicated science benchmarks are unavailable.

Rankings

Domain#RankScoreSource
Agents & Tools46
58.0
LS
Code Ranking210
39.0
AA
General Ranking333
33.0
AA
Multimodal Ranking15
87.0
LS
Reasoning38
74.0
LS
Science214
46.0
AA

Benchmark Scores (LLM Stats)

Agents

OSWorld-Verified72.1%SR
Terminal-Bench 2.060.0%SR
MCP Atlas57.7%SR
SWE-Bench Pro54.4%SR
Toolathlon42.9%SR

Biology

GPQA88.0%SR

Communication

Tau2 Telecom93.4%SR

General

MMMU-Pro76.6%SR
MRCR v2 (8-needle)33.6%SR

Math

Humanity's Last Exam28.2%SR

Multimodal

OmniDocBench 1.587.4%SR

Reasoning

Graphwalks BFS <128k76.3%SR
Graphwalks parents <128k71.5%SR

AA Evaluation Indices

Coding Index
25.3
Intelligence Index
23.3
Gpqa
0.6
Scicode
0.4
Ifbench
0.4
Lcr
0.3
Tau2
0.2
Terminalbench Hard
0.2
Hle
0.1

LLM Stats Category Scores

Structured Output
90
Biology
90
Chemistry
90
Communication
90
Physics
90
Multimodal
80
Spatial Reasoning
70
Vision
70
General
70
Tool Calling
60
Agents
60
Code
60
Reasoning
60
Long Context
30
Math
30
Healthcare
20

Pricing

Input Price$0.75 / 1M tokens
Output Price$4.5 / 1M tokens
Blended Price (3:1)$1.688 / 1M tokens

Speed

Tokens/sec157.4 tokens/s
Time to First Token0.56s
Time to Answer0.56s

Available Providers

(LS internal units)
ProviderInput PriceOutput Price
OpenAI750K4.5M

External Sources